Seasonal changes in open water area in the vicinity of penguin colonies near Palmer Station, northwestern Antarctic Peninsula

The sea ice season in the western Antarctic Peninsula (WAP) region shows some of the fastest rates of change observed globally since 1979.  The Palmer Long-Term Ecological Research (PAL LTER) program, focused on the WAP since 1991, has documented changes both in the ocean (physical and biogeochemical) and at every trophic level. Although co-varying changes are evident, the mechanisms linking environmental change to biological response remain largely elusive.  Part of the problem has been a lack of year-round in situ data, though this is beginning to change with the addition of autonomous sampling (e.g., ocean moorings and gliders).  Another problem is that the readily available (satellite-derived) sea ice record is of coarse spatial resolution (25 km²), appropriate for characterizing regional to global changes in sea ice cover, but not so appropriate for resolving variability in open water fraction, or for relating sea ice variability to penguin foraging behavior.  In winter, open water areas (within the pack ice) are associated with excessively large air-sea gas and heat exchanges and are important access points for penguins, seals and whales.  In early spring, open water areas become a hotbed of activity as phytoplankton begin to bloom within the sea ice and open water areas, attracting high concentrations of marine life.  Within this context we seek to resolve the seasonal changes in open water area in the near vicinity of Palmer Station, where PAL has been sampling twice weekly over spring-summer since 1993, and where records on seabirds go back to the early 1970s.  To help resolve these changes, we examine high-resolution satellite imagery made available by the Polar Geospatial Center (http://www.pgc.umn.edu/).  The high-resolution (mostly optical) imagery, available back to ~2006, is examined in conjunction with the coarser resolution passive microwave data.  The latter provide daily time series and are used to help interpolate the information gained from the high-resolution data (which are less frequently available due to clouds or low/no light conditions).